Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
Still an important development. Anyone knows what problem this chip is designed to solve?
in the article it said it is the multiparticle quantum walk problem. From my understanding it is solving the question of how more than 1 quantum particles would move in a random walk.
Ooh so more a optical quantum asic, still a big step it might mean they can start using the theory to R&D real hardware with it.
Plz no more Dylan Patel, he's the semiconductor Gordon Chang.
Today, I was digging the lithograph optics. What I found is quite intimidating. I found the specs for future high-NA EUV optics.
The maximum surface roughness is 120 picometers. The average roughness is around 20 picometers (0.02 nanometers). This is around 84 times more accurate than the mirrors found on the Hubble Space Telescope and 333 times more accurate than the mirrors on visible-light telescopes. It is 6 times more accurate than the lenses in the scientific instruments in particle accelerators. (I had to research for these figures too.) So we can conclude that future EUV optics are at whole another league when it comes to accuracy. Then I looked at methods for polishing extremely accurate optics. I found an article from 2009.
The most accurate method here is elastic emission machining. It was achieving sub-100 picometers (0.1 nanometers) average roughness in 2009. It achieves it without adding any material to the surface or causing any chemical reactions. It is a purely mechanical method. Also, that level of accuracy is the accuracy of the current EUV lithograph optics. When I started reading on it, I found that it was indeed the method being used for current EUV optics.
What I cannot find is if China has this technology. Can some of you guys learn it?
The maximum surface roughness is 120 picometers. The average roughness is around 20 picometers (0.02 nanometers). This is around 84 times more accurate than the mirrors found on the Hubble Space Telescope and 333 times more accurate than the mirrors on visible-light telescopes. It is 6 times more accurate than the lenses in the scientific instruments in particle accelerators. (I had to research for these figures too.) So we can conclude that future EUV optics are at whole another league when it comes to accuracy. Then I looked at methods for polishing extremely accurate optics. I found an article from 2009.
The most accurate method here is elastic emission machining. It was achieving sub-100 picometers (0.1 nanometers) average roughness in 2009. It achieves it without adding any material to the surface or causing any chemical reactions. It is a purely mechanical method. Also, that level of accuracy is the accuracy of the current EUV lithograph optics. When I started reading on it, I found that it was indeed the method being used for current EUV optics.
What I cannot find is if China has this technology. Can some of you guys learn it?
EUV usually uses multilayer mirrors of Mo/Si thin films. Key word is dielectric mirror. You don't need to machine too much, it is all from vapor deposition. You just need an atomically flat substrate (regular silicon wafers) and a self limiting coating method like atomic layer deposition.
Silicon wafers are typically good enough as a substrate, as they are already atomically flat:
I think that EUV is going to have rapidly diminishing returns. Costs are going to skyrocket due to the ridiculous machining tolerances involved. 120 picometers is literally the bond length of a single Si-Si bond. That is insane. It isn't even measurable especially across an entire wafer (you can measure it at a tiny spot using AFM or STM but field of view is microns) so how do they know how flat it is??
I think the Chinese have working on this using Ion Beam Figuring to reach sub-nanometer polishing but all the papers i have seen are just experimental.
- Status